Scheduling and path planning collaborative optimization method for mobile robot fulfillment system

Abstract

The invention discloses a scheduling and path planning collaborative optimization method for a mobile robot fulfillment system, and belongs to the technical field of intelligent warehousing, and the method comprises the steps: calculating a conflict-free path for carrying AGVs and a chromosome fitness value corresponding to the conflict-free path for carrying the AGVs in consideration of the condition of heterogeneous AGV collaborative operation in warehousing operation; performing a selection operation, a multi-position crossover operation and a single-point mutation operation on the current population based on the chromosome fitness value until an obtained new-generation population reaches a termination condition, and selecting an optimal AGV scheduling scheme and a path planning result from the new-generation population; the scheduling and path planning of the carrying AGV and the cleaning AGV in the mobile robot fulfillment system are optimized at the same time by adopting a double-layer heuristic algorithm, and compared with a traditional single optimization method only facing the carrying AGV, the collaborative optimization of heterogeneous AGV scheduling can be realized, the method is more practical, and the scheduling and planning efficiency is higher.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of robot control, and more particularly relates to a scheduling and path planning collaborative optimization method for a mobile robot fulfillment system. BACKGROUND

[0002] With the development of intelligent warehousing and automated logistics systems, mobile robot fulfillment systems have been widely applied in complex operation environments such as e-commerce and retail warehouses. In such systems, AGVs undertake various tasks such as shelf carrying, order delivery, and cleaning and maintenance. Traditional scheduling and path planning methods focus on single type carrying AGVs, and improve operation efficiency through task allocation and path obstacle avoidance.

[0003] However, in actual fulfillment systems, there are often multiple heterogeneous AGVs working collaboratively, for example, carrying AGVs are responsible for shelf carrying tasks, and cleaning AGVs are responsible for passageway cleaning tasks. Such heterogeneous structure brings system flexibility, but also introduces complex scheduling problems such as path conflict, resource coordination, and task timing dependency. Currently, there is a lack of a unified modeling and collaborative optimization method for multi-heterogeneous AGV task scheduling and path planning, which affects system operation efficiency. SUMMARY

[0004] In view of the above defects or improvement needs of the prior art, the present application provides a scheduling and path planning collaborative optimization method for a mobile robot fulfillment system, which aims to solve the technical problem that the robot fulfillment system in the prior art cannot simultaneously optimize the scheduling and path planning of multiple robots.

[0005] To achieve the above-mentioned purpose, according to one aspect of the present application, a scheduling and path planning collaborative optimization method for a mobile robot fulfillment system is provided, comprising: S1: obtaining the working task information and state position information of carrying AGVs and cleaning AGVs in the mobile robot fulfillment system to initialize a population and enter S2; wherein the mobile robot fulfillment system comprises multiple carrying AGVs, multiple cleaning AGVs, multiple movable shelves, and multiple picking workstations; S2: performing path planning on the AGV scheduling scheme in the current population to obtain a conflict-free path for the carrying AGV, and entering S3; S3: based on the conflict-free path of the carrying AGV, performing path planning for the cleaning AGV to obtain a conflict-free path for the cleaning AGV, the conflict-free paths of the carrying AGV and the cleaning AGV both avoid obstacles in the road section and have no conflict section; S4: calculating the fitness value of the chromosome corresponding to the conflict-free path of the carrying AGV and the conflict-free path of the carrying AGV; S5: performing a selection operation, a multi-position crossover operation and a single-point mutation operation on the current population based on the chromosome fitness value to obtain a new generation population; S6: determining whether the new generation population reaches a termination condition; if yes, selecting an optimal AGV scheduling scheme and path planning result from the new generation population, and if no, returning to S2.

[0006] Further, the S3 comprises: S31: uniformly grid dividing a warehouse operation area, and assigning a corresponding cleaning AGV to each grid; S32: obtaining a global time window table of nodes corresponding to each grid according to the scheduling scheme and the path planning result of the carrying AGV, and recording the occupation state of each grid node at each time point; S33: generating a set of nodes to be cleaned corresponding to each cleaning AGV according to its current position; if the current node of the cleaning AGV is already included in the set of nodes to be cleaned, it is removed from the set of nodes to be cleaned; S34: generating a conflict-free path for the cleaning AGV according to each node in the set of nodes to be cleaned.

[0007] Further, the S34 comprises: S341: selecting an unvisited node closest to the current node from the set of nodes to be cleaned as a target node; S342: planning a conflict-free path segment for the cleaning AGV from the current position to the target node in combination with the global time window table; S343: appending the path segment to the complete path sequence of the cleaning AGV, removing the target node from the set of nodes to be cleaned, taking the target node as a new current node, and returning to S341 until the set of nodes to be cleaned is empty, and then taking the current complete path sequence as the conflict-free path of the cleaning AGV.

[0008] Further, the S342 comprises: selecting part of nodes from the idle state nodes to plan a conflict-free path segment for the cleaning AGV from the current position to the target node by using the A* algorithm with time window in combination with the occupation state of each grid node at each time point in the global time window table.

[0009] Further, the AGV scheduling scheme comprises a carrying AGV scheduling scheme and a cleaning AGV scheduling scheme; and the operation principle of the mobile robot fulfillment system comprises: The channel allows the carrying AGV and the cleaning AGV to travel in both directions; The carrying AGV is allowed to pass through the bottom of the shelf in an empty state; The cleaning AGV only performs periodic operation in the passage area.

[0010] Further, the carrying AGV scheduling scheme comprises: after receiving an order task, starting from a current position to a storage node where a target shelf is located, after completing lifting operation of the target shelf, carrying the target shelf to a designated order picking station for order picking; after order picking is completed, carrying the shelf back to the original storage position.

[0011] Further, the cleaning AGV scheduling scheme comprises: traversing the passage node or edge according to a planned path in a set period to complete the covering ground cleaning task, and avoiding path conflict with the carrying AGV.

[0012] According to another aspect of the present application, a scheduling and path planning collaborative optimization device of a mobile robot fulfillment system is provided, comprising: an acquisition module, a first planning module, a second planning module, a calculation module, an update module and an iteration module; The acquisition module is configured to acquire working task information and state position information of the carrying AGV and the cleaning AGV in the mobile robot fulfillment system, to initialize a population, and enter the first planning module; wherein the mobile robot fulfillment system comprises a plurality of carrying AGVs, a plurality of cleaning AGVs, a plurality of movable shelves and a plurality of order picking stations; The first planning module is configured to perform path planning on the AGV scheduling scheme in the current population to obtain a non-conflict path of the carrying AGV, and enter the second planning module; The second planning module is configured to perform path planning on the cleaning AGV based on the non-conflict path of the carrying AGV to obtain a non-conflict path of the cleaning AGV, wherein the non-conflict paths of the carrying AGV and the cleaning AGV both avoid obstacles in the path segment and have no conflict path segment; The calculation module is configured to calculate the non-conflict path of the carrying AGV and a chromosome fitness value corresponding to the non-conflict path of the carrying AGV; The update module is configured to perform selection operation, multi-position crossover operation and single-point mutation operation on the current population based on the chromosome fitness value to obtain a new generation population; The iteration module is configured to judge whether the new generation population reaches a termination condition; if yes, select an optimal AGV scheduling scheme and path planning result from the new generation population; if not, return to the first planning module.

[0013] According to another aspect of the present application, a control system of a mobile robot fulfillment system is provided, comprising a memory and a processor, the memory stores a computer program, and the processor implements the steps of the scheduling and path planning collaborative optimization method of the mobile robot fulfillment system when executing the computer program.

[0014] According to another aspect of the present application, a computer readable storage medium is provided, which stores a computer program, the computer program being executed by a processor to implement the steps of the method for collaborative optimization of scheduling and path planning of a mobile robot fulfillment system.

[0015] Overall, compared with the prior art, the above technical solutions conceived by the present application can achieve the following beneficial effects: (1) The present application provides a method for collaborative optimization of scheduling and path planning of a mobile robot fulfillment system, which considers the collaborative operation of heterogeneous AGVs in warehouse operations, and simultaneously optimizes the scheduling and path planning of both the transport AGVs and the cleaning AGVs in the mobile robot fulfillment system using a double-layer heuristic algorithm. Compared with the traditional single optimization method that only targets transport AGVs, the present application can achieve collaborative optimization of heterogeneous AGV scheduling, is more practical, and has higher scheduling and planning efficiency.

[0016] (2) The present application generates a conflict-free path for the cleaning AGV based on the global time window table and the set of nodes to be cleaned corresponding to the current position of each cleaning AGV, which is simple to operate and has high execution efficiency.

[0017] (3) The present application designs a double-layer heuristic solving algorithm: the upper layer algorithm realizes task allocation through a genetic algorithm, and the lower layer combines an improved A* path search strategy to plan a conflict-free and executable path trajectory for different types of AGVs, thereby coordinating their time and space behavior in the shared warehouse space and ensuring the consistency and safety of task scheduling and path execution. BRIEF DESCRIPTION OF DRAWINGS

[0018] Figure 1 is a flowchart of the method for collaborative optimization of scheduling and path planning of a mobile robot fulfillment system according to the present application, example 1; Figure 2 is a schematic diagram of a mobile robot fulfillment system according to the present application, example 1; Figure 3 is a schematic diagram of a chromosome of a task allocation algorithm according to the present application, example 1; Figure 4 is a schematic diagram of a crossover operation of a task allocation algorithm according to the present application, example 1; Figure 5 is a schematic diagram of a mutation operation of a task allocation algorithm according to the present application, example 1; Figure 6 is a flowchart of the algorithm according to the present application, example 1. DETAILED DESCRIPTION

[0019] In order to make the purpose, technical scheme and advantages of the present application more clear, the present application is further described in detail below in combination with the drawings and examples. It should be understood that the specific examples described herein are only used to explain the present application and do not limit the present application. In addition, the technical features involved in each embodiment of the present application described below can be combined with each other as long as they do not conflict with each other.

[0020] Embodiment 1 The present embodiment provides a scheduling and path planning cooperative optimization method of a mobile robot fulfillment system, as shown in FIG. 1, which comprises steps S1-S6. Figure 1

[0021] S1: Obtain the work task information and state position information of the carrying AGV and the cleaning AGV in the mobile robot fulfillment system, to initialize the population, and enter S2; wherein the mobile robot fulfillment system comprises a plurality of carrying AGVs, a plurality of cleaning AGVs, a plurality of movable shelves and a plurality of picking workstations.

[0022] Further, the AGV scheduling scheme includes a carrying AGV scheduling scheme and a cleaning AGV scheduling scheme; the operation principle of the mobile robot fulfillment system includes: the passageway allows the carrying AGV and the cleaning AGV to travel in both directions; the carrying AGV is allowed to pass through the bottom of the shelf in the empty state; the cleaning AGV only performs periodic work in the passageway area.

[0023] Specifically, the mobile robot fulfillment system (RMFS system) includes a heterogeneous AGV job system environment model of carrying AGVs and cleaning AGVs, and the environment is defined as follows: (1) The AGVs of various types in the RMFS system run in the warehouse environment, and the passageway allows the AGVs to travel in both directions; (2) The system contains a plurality of AGVs, a plurality of movable shelves and a plurality of picking workstations, wherein the AGVs include two types of carrying AGVs and cleaning AGVs; (3) The carrying AGV can pass through the bottom of the shelf in the empty state; the cleaning AGV only performs periodic work in the passageway area.

[0024] Further, the carrying AGV scheduling scheme includes: after receiving an order task, starting from the current position to the storage node where the target shelf is located, completing the lifting work of the target shelf, and then carrying it to the designated picking workstation for order picking; after the order picking is completed, the shelf is carried back to the original storage position.

[0025] Further, the cleaning AGV scheduling scheme includes: traversing the passageway nodes or edges according to the planned path within a set period to complete the coverage ground cleaning task and avoid path conflicts with the carrying AGV.

[0026] ​Specifically, in the mobile robot fulfillment system, the present application studies the multi-heterogeneous AGV scheduling and path coordination optimization problem, which aims to realize the joint optimization of task allocation and path planning of different types of AGVs. The system contains two types of AGVs that work together: carrying AGVs and cleaning AGVs. Among them, the carrying AGVs are responsible for the carrying operation of the shelves, and the cleaning AGVs periodically perform the ground cleaning of the warehouse passage area. The layout of the mobile robot fulfillment system is shown in Figure 2 as the bird's eye view of the RMFS, which shows the carrying area, the passage area, the storage node and the picking workstation, etc. The parameters involved in S1 are shown in Table 1.

[0027] Table 1 Parameter list

[0028] Since genetic algorithm has strong global search ability and is suitable for solving complex combinatorial optimization problems, it is used for task-AGV mapping optimization in this scheme. The specific design is as follows: Chromosome design: adopt double-layer chromosome design, the first layer is task sequence, and the second layer is AGV sequence. Integer coding is adopted, each gene in the task sequence represents task number, and each gene in the AGV sequence represents AGV number. As shown in Figure 3 , if there are 8 tasks, their numbers are 1-8, and 4 AGVs, their numbers are 1-4. The task execution order is 1-3-4-5-7-6-8, and the AGV sequence corresponding to the task is 1-1-3-2-2-4-3-4.

[0029] Population initialization: generate the upper layer chromosome by randomly sorting the task sequence layer of the chromosome; generate the lower layer chromosome by randomly generating gene bits in the AGV sequence layer, but ensure that each AGV number is generated at least once.

[0030] S2: path planning for the AGV scheduling scheme in the current population to obtain the conflict-free path of the carrying AGV, and enter S3. The steps are as follows: S2, adopt A* algorithm with time window to generate conflict-free path for carrying AGV; in this scheme, a double-layer heuristic path planning algorithm for multi-heterogeneous AGV path planning is proposed, which combines improved A* algorithm to solve the path coordination problem between carrying and cleaning AGVs.

[0031] (1) High-priority path planning layer: for carrying AGV, adopt A* algorithm with time window, preferentially plan carrying task path, and ensure that its path is not disturbed by other AGVs.

[0032] (2) Low-priority path planning layer: for cleaning AGV, adopt coverage path planning, realize comprehensive coverage cleaning of warehouse area under the premise of respecting the space-time constraints of carrying AGV.

[0033] To support the above path planning mechanism, the following key control structures are set up: (1) Global Time Window Table (GTW): records the time occupation of all AGVs at each node, used for conflict detection and avoidance strategy; (2) Conflict detection mechanism: clean AGV detects the space-time conflict with carrying AGV in real time during path planning according to GTW, and selects to detour or wait if conflict occurs; (3) Clean path generation strategy: select uncovered areas one by one through greedy strategy to realize efficient traversal of task area.

[0034] The steps of the A* algorithm with time window used in S2 are as follows: (1) According to the scheduling chromosome output by the upper genetic algorithm, get the starting node nk and task target node nf of the carrying AGV; (2) Initialize the open list open_list and closed list, and add nk to the open list; (3) Select the node with the minimum evaluation function f = g + h in open_list as the current node, where g is the actual cumulative distance from the starting node nk to the current node, and h is the Manhattan distance from the current node to the target node nf; (4) If current == nf, backtrack the path and output the carrying path rd k ; (5) Otherwise, traverse all neighbor nodes of current, for each: a, calculate the current time ct and path cost g to reach neighbor; b, if neighbor is occupied by GTW in the time interval [t, t') and ct ∈ [t, t'), postpone ct to t', and calculate the waiting time; c, update the cost function f, and add the legal neighbor to open_list; (6) Repeat the above steps until a conflict-free path to the target node is found.

[0035] The pseudo code of the A* algorithm with time window is shown in Table 2: Table 2 A* algorithm with time window

[0036] S3: conflict-free path based on the carrying AGV, path planning is performed for the cleaning AGV to obtain a conflict-free path of the cleaning AGV, and the conflict-free paths of the cleaning AGV and the carrying AGV both avoid the obstacles in the road section and have no conflict section.

[0037] Further, S3 includes: S31: uniformly grid dividing the warehouse operation area, and assigning corresponding cleaning AGVs to all grids; S32: obtaining a global time window table of nodes corresponding to each grid according to the scheduling scheme and the path planning result of the carrying AGV, and recording the occupation state of each grid node at each time; S33: generating a corresponding to-be-cleaned node set of each cleaning AGV according to the current position of the cleaning AGV; if the current node of the cleaning AGV is already included in the to-be-cleaned node set, the current node is removed from the to-be-cleaned node set; and S34: generating a conflict-free path of the cleaning AGV according to each node in the to-be-cleaned node set.

[0038] Further, S34 includes: S341: selecting one unvisited node closest to the current node from the to-be-cleaned node set as a target node; S342: planning a conflict-free path segment of the cleaning AGV from the current position to the target node in combination with the global time window table; S343: appending the path segment to a complete path sequence of the cleaning AGV, removing the target node from the to-be-cleaned node set, taking the target node as a new current node, and returning to S341 until the to-be-cleaned node set is empty, and then taking the current corresponding complete path sequence as the conflict-free path of the cleaning AGV. As an optional implementation, S342 includes: selecting part of nodes from idle state nodes to plan a conflict-free path segment of the cleaning AGV from the current position to the target node by using the A* algorithm with time window in combination with the occupation state of each grid node at each time in the global time window table.

[0039] As an optional implementation, the whole step S3 includes the following steps.

[0040] (1) According to the total number of nodes in the warehouse operation area, the nodes are evenly divided into a plurality of sub-areas, and each sub-area is assigned to a corresponding cleaning AGV.

[0041] (2) According to the upper scheduling result, a global time window table is obtained to record the occupation state of each node at each time.

[0042] (3) According to the current position of each cleaning AGV, a corresponding to-be-cleaned node set of the cleaning AGV is generated.

[0043] (4) If the current node is already included in the to-be-cleaned node set, the current node is removed from the set.

[0044] (5) Under the condition that the to-be-cleaned node set is not empty, the following operations are repeatedly performed: a. Select an unvisited node closest to the current node from the set; b. Use the A* algorithm with time window to plan a conflict-free path for the cleaning AGV from the current position to the node in combination with the global time window table; c. Append the path segment to the complete path sequence of the cleaning AGV; d. Remove the node from the set of nodes to be cleaned and update the current node position to the node.

[0045] (6) When all the nodes to be cleaned are visited, output the final cleaning AGV path planning result.

[0046] As an optional implementation, S3, based on the handling path result, uses regional division and greedy strategy to generate a conflict-free path for the cleaning AGV, and the algorithm steps are as follows: (1) According to the total number of nodes in the warehouse operation area |N|, divide it into several sub-regions according to the number, and allocate each sub-region to the corresponding cleaning AGV set K s in the middle; (2) According to the upper layer scheduling result, obtain the global time window table for recording the occupation state of each node at each time; (3) According to the current position current of each cleaning AGV, generate its corresponding set of nodes to be cleaned Nu; (4) If the current node is already included in the set of nodes to be cleaned, remove it from the set; (5) Under the condition that the set of nodes to be cleaned is not empty, the following operations are executed in a loop: a. Select an unvisited node n closest to the current node from the set; b. Use the A* algorithm with time window to plan a conflict-free path rc n for the cleaning AGV from the current position to the node in combination with the global time window table; c. Add the path segment rc n to the complete path sequence rc of the cleaning AGV; d. Remove the node from the set of nodes to be cleaned and update the current node position to the node current = nk; (6) When all the nodes to be cleaned are visited, output the final cleaning AGV path planning result.

[0047] The pseudo code of the cleaning AGV coverage path algorithm used in S3 is shown in Table 3.

[0048] Table 3

[0049] S4: Calculate the conflict-free path of the carrying AGV and the chromosome fitness value corresponding to the conflict-free path of the carrying AGV. In the present scheme S4, the latest global time window table is obtained according to the paths of the carrying AGV and the cleaning AGV, and the cumulative delay time of all tasks of the optimization target is obtained. Since the larger the fitness function is, the greater the probability of the chromosome being selected into the next generation is, the fitness function is set as the inverse of the optimization target. In order to guarantee the timeliness of the carrying task, the optimization target is set to minimize the cumulative delay time of all tasks. Since the larger the fitness function is, the greater the probability of the chromosome being selected into the next generation is, the fitness function is set as the inverse of the optimization target, and the calculation formula is as follows: .

[0050] S5: Select the current population based on the chromosome fitness value, and perform multi-position crossover operation and single-point mutation operation to obtain a new generation of population.

[0051] In the present scheme S5, in order to obtain a better scheduling and path planning scheme, genetic algorithm is used to perform evolution operation on the current population, which specifically includes: (1) Selection operation: according to the fitness value corresponding to each chromosome, a roulette selection strategy is adopted to select individuals with higher fitness value into the next generation in a probabilistic manner, so as to ensure that high-quality solutions have a greater chance of being retained, while maintaining population diversity; Selection operation: roulette selection method is used for parent individual selection, so as to ensure that excellent individuals have a higher probability of reproduction.

[0052] (2) Crossover operation: for the selected individual pair, multi-position crossover operation is adopted, that is, fragment exchange is performed on two chromosomes at multiple preset crossover points to generate new individual solutions, so as to realize information recombination by combining the local excellent structures of different individuals, and enhance the search ability of the population; Crossover operation: multi-position crossover operation is adopted. First, a random number f not exceeding the number of tasks is generated, and then f gene positions are randomly selected in the task sequence layer, and the genes of the selected gene positions in parent 1 are retained. Second, the genes of the unselected gene positions in parent 2 are retained and placed in the corresponding gene positions of the offspring. Third, the genes of the selected gene positions in parent 1 are placed in the corresponding gene positions of the offspring. The specific process is shown in Figure 4 .

[0053] (3) Mutation operation: in order to avoid falling into local optimum, single-point mutation operation is further introduced, that is, a gene position in the chromosome is randomly changed with a certain probability, so as to increase the diversity and global search ability of the population. Mutation operation: unit position mutation operation is adopted. A random number f not exceeding the number of tasks is generated, and then the fth gene position is selected in the AGV sequence layer, and the AGV number corresponding to the fth task is randomly generated. If the AGV number corresponding to the fth task in this position exists only one in the AGV sequence, the constraint that each AGV is assigned is kept unchanged. The specific process is shown inFigure 5 as shown.

[0054] S6: judging whether the new generation population reaches the termination condition; if yes, selecting the optimal AGV scheduling scheme and path planning result from the new generation population, and if no, returning to S2. That is, in the scheme S6, it is judged whether the termination condition is reached, if yes, the AGV scheduling and path result are output, and if no, the next round of iterative evolution is returned to S2.

[0055] The present application adopts a double-layer heuristic algorithm to schedule and cooperatively optimize the heterogeneous AGV system. Specifically, the upper layer uses a genetic algorithm to complete the task allocation scheme of the carrying AGV and the cleaning AGV, and the lower layer generates the conflict-free execution path of each type of AGV in the shared space based on an improved A* path search algorithm. In the optimization process, the task allocation result is used as the input of the path planning module to generate the corresponding heterogeneous AGV path scheme, and the feasibility and merits and demerits of the current task allocation scheme are evaluated according to the path execution feedback. Through the above cooperative running mechanism of the upper layer scheduling and the lower layer path, the population evolution iteration is continuously carried out until the termination condition is met, and finally the cooperative operation scheme that meets the scheduling constraints and has the optimal path is output. The overall flowchart of the algorithm is as shown. Figure 6 as shown.

[0056] Embodiment 2 The present embodiment provides a scheduling and path planning cooperative optimization device of a mobile robot fulfillment system, comprising an acquisition module, a first planning module, a second planning module, a calculation module, an updating module and an iteration module.

[0057] The acquisition module is used to acquire the working task information and state position information of the carrying AGV and the cleaning AGV in the mobile robot fulfillment system, to initialize the population, and to enter the first planning module; wherein the mobile robot fulfillment system comprises multiple carrying AGVs, multiple cleaning AGVs, multiple movable shelves and multiple picking workstations.

[0058] The first planning module is used to perform path planning on the AGV scheduling scheme in the current population to obtain the conflict-free path of the carrying AGV, and to enter the second planning module.

[0059] The second planning module is used to perform path planning for the cleaning AGV based on the conflict-free path of the carrying AGV to obtain the conflict-free path of the cleaning AGV, and the conflict-free paths of the carrying AGV and the cleaning AGV both avoid the obstacles in the road section and have no conflict section.

[0060] The calculation module is used to calculate the conflict-free path of the carrying AGV and the fitness value of the chromosome corresponding to the conflict-free path of the carrying AGV.

[0061] An updating module is configured to perform selection operation, multi-position crossover operation and single-point mutation operation on the current population based on the chromosome fitness value to obtain a new generation population.

[0062] An iteration module is configured to determine whether the new generation population reaches a termination condition; if yes, select an optimal AGV scheduling scheme and path planning result from the new generation population; if no, return to the first planning module.

[0063] Embodiment 3 The embodiment provides a control system of a mobile robot fulfillment system, including a memory and a processor, the memory stores a computer program, and the processor implements the steps of the scheduling and path planning collaborative optimization method of the mobile robot fulfillment system when executing the computer program.

[0064] Embodiment 4 The application also relates to a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to implement the steps of the method.

[0065] Specifically, the memory can include a high-speed random access memory, and can also include a non-volatile memory, for example, a hard disk, a memory, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, at least one magnetic disk storage device, a flash memory device, or other volatile solid-state memory devices.

[0066] Embodiment 5 The embodiment of the application provides a computer program product or a computer program, which comprises computer instructions stored in a computer readable storage medium. A processor of a computer device reads the computer instructions from the computer readable storage medium, and the processor executes the computer instructions, so that the computer device executes the steps of the method of the above embodiment of the application.

[0067] The technical features of the above embodiments can be combined in any manner, and to make the description concise, all possible combinations of the technical features in the above embodiments are not described, however, as long as the combinations of the technical features do not exist contradictions, they should be considered as the scope of the present application. It should be noted that the "in an embodiment of the present application", "for example", "for example" and the like in the present application are intended to illustrate the present application, and are not used to limit the present application.

[0068] The above embodiments only express several implementation manners of the present application, and the description is more specific and detailed, but it should not be understood as a limitation to the scope of the application. It should be noted that for those skilled in the art, without departing from the concept of the present application, several modifications and improvements can be made, which are all within the protection scope of the present application.

Claims

1. A method for coordinated optimization of scheduling and path planning in a mobile robot execution system, characterized in that, include: S1: Obtain the work task information and status location information of the transport AGV and cleaning AGV in the mobile robot execution system to initialize the population and proceed to S2; wherein, the mobile robot execution system includes multiple transport AGVs, multiple cleaning AGVs, multiple mobile shelves and multiple picking workstations; S2: Perform path planning on the AGV scheduling scheme in the current population to obtain a conflict-free path for the transport AGVs, and then proceed to S3; S3: Based on the conflict-free path of the transport AGV, a path planning is performed for the cleaning AGV to obtain a conflict-free path for the cleaning AGV. Both the conflict-free path of the cleaning AGV and the conflict-free path of the transport AGV avoid obstacles in the road segment and there are no conflicting road segments between them. S4: Calculate the conflict-free path of the transport AGV and the chromosome fitness value corresponding to the conflict-free path of the transport AGV; S5: Based on the chromosome fitness value, perform selection, multi-position crossover, and single-point mutation operations on the current population to obtain a new generation population; S6: Determine whether the new generation of the population has reached the termination condition; if yes, select the optimal AGV scheduling scheme and path planning result from the new generation of the population; otherwise, return to S2.

2. The scheduling and path planning collaborative optimization method for a mobile robot execution system as described in claim 1, characterized in that, S3 includes: S31: Divide the warehouse operation area into uniform grids and assign corresponding cleaning AGVs to all grids; S32: Based on the scheduling scheme and path planning results of the transport AGV, obtain the global time window table of the nodes corresponding to each grid, which is used to record the occupancy status of each grid node at each time. S33: Generate a set of nodes to be cleaned for each cleaning AGV based on its current position; if the current node of the cleaning AGV is already included in the set of nodes to be cleaned, remove it from the set of nodes to be cleaned. S34: Generate a conflict-free path for the cleaning AGV based on each node in the set of nodes to be cleaned.

3. The scheduling and path planning collaborative optimization method for a mobile robot execution system as described in claim 2, characterized in that, S34 includes: S341: Select the nearest unvisited node from the set of nodes to be cleaned as the target node; S342: Combine the global time window table to plan a conflict-free path segment from the current position to the target node for the cleaning AGV; S343: Append the path segment to the complete path sequence of the cleaning AGV, remove the target node from the set of nodes to be cleaned, take the target node as the new current node, and return to S341 until the set of nodes to be cleaned is empty. Then, take the current corresponding complete path sequence as the conflict-free path of the cleaning AGV.

4. The scheduling and path planning collaborative optimization method for a mobile robot execution system as described in claim 3, characterized in that, S342 includes: By combining the occupancy status of each grid node in the global time window table at each time, the A* algorithm with time window is used to select some nodes from the idle nodes to plan a conflict-free path segment from the current position to the target node for the cleaning AGV.

5. The scheduling and path planning collaborative optimization method for a mobile robot execution system as described in claim 1, characterized in that, The AGV scheduling scheme includes a material handling AGV scheduling scheme and a cleaning AGV scheduling scheme; the operating principles of the mobile robot performance system include: The passageway allows the transport AGV and the cleaning AGV to travel in both directions; The transport AGV is allowed to travel under the shelf when unloaded; The cleaning AGV only performs periodic operations in the passageway area.

6. The scheduling and path planning collaborative optimization method for a mobile robot execution system as described in claim 5, characterized in that, The AGV dispatching scheme includes: after receiving an order task, starting from the current location, proceeding to the storage node where the target shelf is located, lifting the target shelf, and then transporting it to the designated picking workstation for order picking; after the order picking is completed, the shelf is then transported back to its original storage location.

7. The scheduling and path planning collaborative optimization method for a mobile robot execution system as described in claim 5, characterized in that, The cleaning AGV scheduling scheme includes: traversing channel nodes or edges according to a planned path within a set period to complete the comprehensive ground cleaning task and avoid path conflicts with transport AGVs.

8. A scheduling and path planning collaborative optimization device for a mobile robot execution system, characterized in that, include: The module includes an acquisition module, a first planning module, a second planning module, a calculation module, an update module, and an iteration module. The acquisition module is used to acquire the work task information and status location information of the transport AGV and the cleaning AGV in the mobile robot performance system, so as to initialize the population and enter the first planning module; wherein, the mobile robot performance system includes multiple transport AGVs, multiple cleaning AGVs, multiple mobile shelves and multiple picking workstations; The first planning module is used to perform path planning on the AGV scheduling scheme in the current population to obtain a conflict-free path for the transport AGV, and then enter the second planning module; The second planning module is used to perform path planning for the cleaning AGV based on the conflict-free path of the transport AGV to obtain a conflict-free path for the cleaning AGV. Both the conflict-free path of the cleaning AGV and the conflict-free path of the transport AGV avoid obstacles in the road segment and there are no conflicting road segments between them. The calculation module is used to calculate the conflict-free path of the transport AGV and the chromosome fitness value corresponding to the conflict-free path of the transport AGV. The update module is used to perform selection, multi-position crossover, and single-point mutation operations on the current population based on the chromosome fitness value to obtain a new generation population. The iteration module is used to determine whether the new generation population has reached the termination condition; if so, the optimal AGV scheduling scheme and path planning result are selected from the new generation population; if not, the process returns to the first planning module.

9. A control system for a mobile robot execution system, comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 7.

Images